A Primer on Benefit Design

I recently finished a really excellent course on Healthcare Economics and wanted to reflect on my favorite unit in the course: Benefit Design. In order to lay the groundwork for what Benefit Design is and why it is important, I'll first provide some general background on how health insurance works in the United States. Then, I'll discuss the current state of the US health care system in terms of health and spending which will help set the stage for the role of Benefit Design. Finally, I'll dig into what Benefit Design is and look at some standard Benefit Design tools.

Overview of Health Insurance

Health insurance in the US is both complex and unique, relative to other countries. The simple explanation of how it all works is there are three primary actors: the patient, the provider, and the payer. The patient is you, the individual, the person paying for and receiving health care. The provider is the entity that delivers health care services to you, typically a primary care doctor or a specialist. And the payer is the entity that pays the provider for your health care.

Most working adults pay for and receive health insurance through their employer, so called "employer-sponsored" health insurance. An employer offers their employees a list of insurance plans they can choose from, and that plan dictates the premiums (which are jointly paid by employers and employees to the payers in exchange for health insurance) that are paid to the payer every month in exchange for coverage of the costs of health care services as the need arises. The details of what services are covered, and to what extent, are governed by the health insurance plan.

It may be surprising for US readers to learn that tying health insurance to employment status is not a standard practice across the world. While not the focus of this post, the origin of employer-sponsored health insurance is interesting, and serves as a compelling example of how current policies often have surprising roots. During World War II, many people were being shipped overseas which led to a labor shortage in the US.  The diminished supply of workers caused companies to want to offer higher wages. The Roosevelt administration wisely worried that this would lead to inflation, so they instituted a wage cap. Employer-sponsored health insurance then took off as a benefit that companies could offer to prospective employees in lieu of higher wages, and it become entrenched when Congress ruled that the amount that your employers pays for your premiums is not considered taxable income. This meant that, from an employees perspective, it was substantially cheaper to obtain health insurance through your job because your employer can subsidize health insurance premiums with pretax dollars. This coupling between employment and health insurance was meant to serve as an incentive to work during wartime. Without getting too sidetracked as to whether employer-sponsored health insurance with a tax exemption makes sense in today's world, at the very least it serves as a compelling example of how temporary circumstances can give rise to policies that endure long after those circumstances cease to exist.

Current State of US Health Care

You don't need to take a healthcare economics course to be aware that things aren't working as they should be in US health care. This report shows that relative to other OECD countries, we seem to lead and lag in all the wrong metrics:

• We pay the greatest percentage of our GDP to health care (almost double the average of other OECD countries).
• Life expectancy in the US is at least three years lower than OECD average, and declining.
• We have the highest rates of infant and maternal deaths, high rates of death from COVID-19, highest rates of chronic disease, and highest obesity rate of 40% (!).\
  

We spend trillions of dollars on health care, which is honestly an incomprehensible number. But it's important to note that it's not just abstract systemic costs that are rising. The amount that individual patients pay out of pocket has actually been rising the fastest (faster than system-wide health spending, wages, and inflation), per this chart from the course:

and on top of that, health insurance premiums are set to increase by 6.5% this year, the biggest increase in more than a decade (and mostly driven by the availability of costly new weight loss drugs like Ozempic and Wegovy). Most businesses, and surely most employees, are ill-prepared to absorb a substantial increase in healthcare costs, especially if the economy tips into a recession.

Our rising costs paying for diminished health outcomes is, at this point, a common narrative that many are aware of. While it is a dire state of affairs, I do want to offer a couple caveats that I don't believe are mentioned enough:

• The US is the leader in medical technology innovation, and a lot of our high and growing health care spend funds the development of new drugs, vaccines, and other types of medical technology that the rest of the world benefits from (the COVID-19 vaccines being the obvious, recent example).
• It's well known that health status and socio-economic status are tightly coupled, and the US has a vastly different socio-economic environment than other OECD countries due to immigration, income inequality, etc.
  

It's easy to look at these numbers and feel that the scope and scale of the problems with US health care are insurmountable, but I'm encouraged by the fact that there is a lot of opportunity for improvement. One study estimates that approximately 25% of total health care spending is due to waste. The course distilled the causes of our out of control health care spending to three main things:

• Overuse - consuming too much low quality care
• Underuse - not consuming enough high quality care
• High prices
  

The encouraging thing is that individual patients can help to alleviate the effects of each of these causes by making better decisions about their own care. Regarding overuse, for example, one study found that, at one point, 42% of Medicare beneficiaries received a low-value service (incidentally, there are various initiatives devoted to categorizing the value of health care services such as ChooseWisely) which resulted in $8.5 billion of unnecessary healthcare spending. There are a few intuitive causes for overuse. From a patient's perspective, if someone else (the payer) is footing all or most of the bill, why not consume additional health care services (such as labs, diagnostics, CT scans, etc) just to be safe? The tendency of an individual to change their behavior or risk tolerance when they don't bear the financial costs is called moral hazard. The provider also has an incentive to induce overuse of tests, scans, and the like to defend themselves against potential medical malpractice lawsuits. They want to be as defensive and safe as possible so they don't become the doctor that lazily failed to catch some extremely costly or threatening condition when it was treatable. The problem is that unnecessary tests and treatments have costs associated with them including anxiety, radiation, unnecessary antibiotics, false positives (which lead to more unnecessary tests), to name a few.

Regarding underuse, the earlier study about health care waste noted that failure to adopt preventive care practices can cost $111 billion per year, and that less than 10% of adults over 35 had received all recommended preventive services. And it's not just lack of prevention, it's also been found that only 70% of people take prescriptions and 55% of people take statins when they're prescribed for high cholesterol. It's not terribly complicated that if we don't do the things that keep us healthy, we pay more down the road when we get sick.

Regarding high prices, there are numerous examples in health care of a service whose price varies widely between providers while quality does not. One paper examined how insured people selected providers for an MRI scan (which does not vary widely in quality) and paper found that on average, patients bypassed six lower-priced providers between their homes and treatment locations and spend almost $84 more out of pocket for the service they eventually used.

I struggle to imagine a scenario outside of healthcare where I would inconvenience myself and pay more for an equivalent or inferior service. The obvious question, then, is why patients would behave in such an obviously irrational way. The paper found that the strongest factor in determining which provider a patient ultimately saw was the recommendation of the referring physician. Most physicians refer patients to practices or facilities they are familiar with, not necessarily to the most cost effective provider. Generally speaking, physicians are held in very high esteem by patients. They are highly credentialed, well paid, and they provide counsel to people when they are at their most vulnerable. I do believe that most physicians want to do good by their patients, but it is critical to realize that they have incentives that don't always align with the physical or financial health of the patient. I'll leave it there as proper treatment of the incentives facing providers is a post all its own.

Benefit Design

In any case, this misalignment of incentives, and the associated waste it creates, is where Benefit Design comes in. Benefit Design aims to create incentives on behalf of the patient to make better decisions regarding their care. The main lever used in the various Benefit Design tools is cost sharing, which refers to the financial responsibility that the patient bears in paying for health care. Higher cost sharing means the patient pays a higher percentage of the costs associated with a given service (and the payer pays a lower percentage).

The most recognizable benefit design tool is also the simplest: High Deductible Health Insurance Plans (HDHPs). A deductible refers to the amount the patient pays out of pocket before their plan starts to pay. The idea behind HDHPs is to raise deductibles so that patients assume more of the risk of their health care. The merits of HDHPs are that they are simple to administer. In their purest form, you basically just increase the deductible component of a health plan, and voila, patients will realize they have skin in the game and start making decisions that are price conscious and clinically informed. In practice, it doesn't quite shake out this way. HDHPs, while well intentioned, have a lot of issues. For one, they are a blunt instrument, and make no distinction between high value and low value care. Also, in practice, it's been found that HDHPs cause patients to cut back on healthcare spending across the board rather than shifting their consumption of health care to only those services that are high value. Secondly, for a lot of patients, HDHPs simply don't make financial sense. This report shows how people at various deductible levels have insufficient or barely sufficient savings to cover the costs of a medical emergency, and are forced to go to extreme measures (such as dipping into retirement savings or spending less on food) to pay for medical bills. So, while the idea of HDHPs is to ensure patients have some money on the line, in practice, many patients with HDHPs are instead risking financial ruin.

Reference pricing is another benefit design tool that offers a more nuanced approach to cost sharing, at the expense of a little complexity and loss of generality. The idea behind reference pricing is for payers to offer a maximum amount that they'll reimbursement for a particular service. So, using our earlier example, if a patient is shopping for an MRI scan, the payer might set a limit of $300 that they'll reimburse for that scan. If a patient obtains the scan for more than $300, they pay the difference. In practice, reference based pricing has worked well on two levels. One, it gives patients an appropriate amount of "skin in the game" by encouraging them to seek out price competitive providers for a particular range of services. Secondly, and likely a byproduct of the first benefit, is that it exerts some healthy downward pressure on market prices. One study found that applying reference pricing to orthopedic surgery in California resulted in more hospitals charging prices below the reference pricing rate. The reference price became the "center of gravity", dragging the prices of more expensive providers down. This is a very good thing because it reduces variation (which, in the context of a market for health care services, is synonymous with inefficiency).  The main downside of reference pricing is that it can only be applied to "shoppable" services. Shoppable services fulfill two criteria: their quality doesn't vary widely, and patients must have ample time to devote to searching for suitable providers. It should not be applied to emergency services.

Yet another benefit design tool is one that most are probably familiar with: tiered networks. The idea behind tiered networks is pretty intuitive. A payer puts providers into tiers based on their quality, and patients who consume care from top tier providers will enjoy more generous cost sharing (less out of their own pocket) than if they had gone to a low tier provider. One study found that the use of tiered networks resulted in a most decrease in aggregate spending for radiology services. The devil, of course, is in the details. How are providers put into tiers? Is it because those providers offer higher quality services than their lower-tier counterparts, or because they're simply more able to handle a higher volume of patients and therefore are able to cut more attractive deals with payers? The details of provider selection are often opaque.

The last benefit design tool I want to mention is referred to as Value-Based Insurance Design (VBID). In my view, it is the most complex benefit tool but also the most nuanced and targeted in its approach. The idea is to offer lower cost sharing for higher value services. Where HDHPs are blunt instruments, VBID takes a "clinically sensitive" approach to cost sharing that varies by service (and, in some VBID implementations, by patient). One question, though, is what is meant by high-value? The answer depends on how VBID is implemented in the context of a health plan. One approach is to not differentiate between patients and just target services that are generally accepted as high value, regardless of the population of patients. Another way of implementing a VBID approach is to target a specific population of patients that have a certain condition (e.g, diabetics) and offer generous reimbursement for specific services. For example, Asheville, North Carolina ran a program that only targeted members who had a preexisting condition of diabetes. The program offered lower copayments for medications that were aimed at managing or treating diabetes (including statins, which were mentioned earlier as one of the high value medications that suffer from only 50% adherence). The results of the program were quite striking, and included "marked increases in medication adherence, a two- to threefold increase in achieving diabetes performance measures, approximately a 50 percent decrease in average annual sick leave".

Conclusion

It is interesting to note that these and other benefit design tools are not mutually exclusive. They are best viewed as complementary to one another and can often be combined, which offers designers of health plans a rich set of options for encouraging patients to make decisions that make financial and clinical sense.

Benefit design on it's own can't fix American health care, as there are still many structural issues in the system (not just the health system but also, for example, the food system, which spills over into health) that make it increasingly difficult for the average American citizen to be healthy. But I was struck by how many of the issues that pertain to health care delivery and consumption can be addressed by rethinking incentives, and I am compelled by the creativity of the solutions that exist in Benefit Design. It's essentially a game in which you try to induce patients to make intelligent clinical and financial decisions, and the limiting factors are the amount of financial risk a patient can bear, as well as the patient's capacity for navigating complexity. I hope you enjoyed this background about benefit design and I'm excited to dig into other topics in health care and economics in future posts.